How Zettlr changed my relation to research literature

After using Zettlr for a couple of years, I have noticed a fundamental change in how I use research literature in my research and writing. In this blog post, I reflect on this development and describe how it happened.

Too little writing, too late

As researchers, we know that publishing our results in peer-reviewed journals and conferences is our research’s primary (if not the only) proof. We researchers should be able to write convincingly in order to present and disseminate our research. Skills and tools for writing –in addition to those for conducting research –are an essential part of our toolbox.

However, partly due to publication bias, writing is often regarded as a separate –preferably short –process that comes after having conducted the research. Popular thinking goes like this: First, you design the research study, then you complete the research (collecting data, analyzing, etc.), and then you write it up and publish the paper.

Writing can, therefore, turn into a burden, something we do as the last step, something that is boring because we already know everything we need to know (we think). Writing gets done out of necessity and, therefore, through suboptimal routines and processes. After all, many of us became researchers not because we wanted to write but to do research. This mindset is witnessed by the myriad of articles and books that try to motivate researchers to write more and write well.

Writing versus writing up

One consequence of this separation of writing as a last “writing up” phase is that we fail to take full advantage of “writing” as a mental and social process of reflection and conceptual clarification. The power of such reflections is not fed into our research process if writing is delayed to a last-minute writing-up phase, when it is too late.

Many scholars of academic writing, therefore, encourage us to make a distinction between “writing” –the reflective and social process of developing our concepts and clarifying our thinking –and “writing up” –i.e., the process of preparing a manuscript for submission to a journal or a conference. Writing –as opposed to writing up –is the process through which we learn throughout our research, in the same way that we learn through data collection and analysis (Mitchell & Clark, 2021). Writing should therefore start on day one and continue throughout the whole research process.

Dealing with research literature

Writing –as opposed to writing up –is particularly important when we deal with research literature. Rich conceptual models in our brains are developed through close interaction with research literature (Leidner & Tona, 2021) and our own first-hand data (Glaser & Strauss, 1967). Research literature gives us a vocabulary to talk about our research and develop a deep understanding of the research phenomenon. In fact, some scholars believe that research is, in reality, a “re-writing” of what other researchers have written before us (Harris, 2017). This is particularly true for qualitative researchers and researchers who work a lot with concepts and conceptual models.

Unfortunately, due to our “writing up” and publication bias, our interaction with research literature is also treated as a “boring” last step in our research. The “state-of-the-art” or “related work” section in our papers is where we have to come up with some references that confirm our line of argument and show the reader that we have a novel contribution to make. Similar to the writing-up phase, this interaction with research literature –and the disciplinary conversations it represents –happens as a final step in our research, without in-depth involvement with what we read.

The publication bias in our tools

The bias towards “writing up” –as opposed to “writing” as part of our research –is not only a part of our processes. It also exists in our digital tools for dealing with literature. As researchers who also write, we have separate tools for writing documents and keeping our references. Our document writing tools are often designed based on the model of a printed document. This is also what our journals and conferences ask for, i.e., a sequence of text followed by a list of references to cited literature. So linear “writing up” thinking is reinforced by our tools and our publishers.

Our references to research literature are often stored in a separate bibliography tool. This tool’s main job –as its name suggests –is to automate creating a bibliography of cited literature for a paper, i.e., the last “References” section in our “written-up” paper. In the name of automation, bibliography tools separate us even more from dealing with research literature. What these tools don’t support well is writing as a reflective dialog with research literature, as discussed above.

I experience that storing my references and PDFs in a bibliography tool such as Zotero creates a barrier between my writing and research literature. Our tools don’t provide a proper way of “going back and forth” between our text and research literature. Before using Zettlr, I tried different approaches –PDF annotations, notes in Zotero, separate summary documents, and even NVivo (a quite advanced qualitative data analysis tool)–to connect my writing to my reading of research literature. These approaches worked somehow, but they resulted in scattered information and, in the long run, continued to isolate my writing from research literature.

Another problem, in my experience, is that both document writing and bibliography tools focus too much on the aesthetics of the text, its formatting into a printable document, and its bibliography style. This means that writing becomes mixed with editing and formatting. This is not a good process, and disrupts the creative process of generating text. The printed document metaphor of writing tools like MS Word additionally discourages one from working with networks of concepts –and the sources of these concepts in the research literature.

Enter Zettlr

Enter the new generation of personal knowledge management tools. These “second brain” tools, such as Zettlr and Obsidian (the two that I have personally used and know quite well) are, from the ground-up, concept-centric as opposed to document-centric tools like MS Word. They encourage methods of information management that try to simulate how we think in networks of concepts. One such method that is quite well-known (and I understand has inspired the creator of Zettlr) is Zettelkasten, attributed to the German sociologist Niklas Luhmann. These tools and methods separate “writing” and “writing up” by focusing on concepts and connections instead of sequential document structures.

However, only a few of these second brain tools are developed for researchers who write. Although these tools are much closer to how we think when generating text, I experienced that one weakness with most of these tools is their lack of support for close integration with research literature. This is where Zettlr excels.

I like to think of Zettlr as an open-source tool for rigorous research. Zettlr is a markdown editor with loads of features for writing research-based text. I don’t intend to provide a beginner’s guide to Zettlr here. Instead I want to focus on how Zettlr has helped me move from a “writing up” mindset to a mindset for “writing” as research. In particular, I want to write about how Zettlr immerses me in research literature while I am writing and allows me to see my writing in the context of this literature.

That said, here is a short description of the features in Zettlr for dealing with references and bibliographies in general (you can skip this list if you already use Zettlr):

  • Zettlr allows you to cite literature in your text. I use Zettlr with Zotero –an open-source tool to deal with references. You can also use a couple of other bibliography tools with Zettlr.
  • Citing Zotero references in Zettlr is similar to how Zotero works in MS Word. You tell the editor you want to cite and search for the reference you want to cite. Once you have found the reference, you enter a citation to it in your text. Simple.
  • When you use Zettlr’s exceptional export functionality, Zettlr will automatically create a properly formatted bibliography section for you at the end of your generated document, along with all your citations where you wanted them. This is also very similar to how Zotero works with MS Word.

Important to mention: Zettlr uses another open-source tool called Pandoc for many of the behind-the-scenes operations that enable the above functionality.

Symbiosis between own writing and research literature

Working with citations and bibliographies is flawless in Zettlr. However, two more subtle features have made a big difference for me when working with research literature.

First, you can right-click on a citation in your text in Zettlr and open the PDF file from your Zotero library directly in your PDF reader. This simple feature has radically reduced the threshold to access research literature while writing. It allows me to effortlessly switch back and forth between my text and my references. As a consequence, the feature has also eliminated the need to write my notes in Zotero (see also next point). If I need to remember what a reference is about I right-click and open the PDF. With proper color-coding in the PDF file I can immediately find the definitions and quotations I need in my writing.

Second, and certainly not unique to Zettlr, you can create, tag, and link text files easily. I used to use Zotero to keep my notes about references. The problem with this approach was that my notes were scattered around in different references and were not easily accessible in aggregated form. More importantly, the notes in Zotero are not interpreted and integrated into my current writing context (Ahrens, 2017) because Zotero (and other bibliography tools) is article-centric and not concept-centric. If I wanted to write about a concept, say digital platforms, it was challenging to integrate all my references that are about digital platforms. In Zettlr, I can easily create a definition note (see next section) called digital platform, and write, cite, and tag all my stuff about digital platforms, including references and other online and offline resources. The thing is that doing this is so low-cost and low-threshold that you don’t hesitate to create new notes for emerging concepts from the literature (Watson & Webster, 2020).

This subtle connection between my writing and my PDFs containing the research literature has had fundamental consequences, at least for me. Combined with the low threshold of creating files in Zettlr, the ability to tag files with keywords, and easy access to related files, Zettlr establishes a symbiosis between my manuscript and the literature. And I love it!

How I use Zettlr to write

Finally, here is a short final section describing how my “Zettelkasten” is organized in Zettlr. I am sure there are better ways to do this, and I would be glad to hear from other qualitative researchers about how you use Zettlr in your research!

In my setup in Zettlr I have created the following note types that somehow resemble the process of research and writing papers:

  • Definition: The most basic note for me is a note that defines a concept. Currently, I have 126 definition notes. Some examples are digital transformation, algorithmic fairness, world-systems theory, generativity, and chatbot. Most of these definition notes use tags and link to other definition notes, creating a conceptual model for my research.
  • Evidence: I use this note type to collect research-based evidence about the connections between concepts. Some examples are use of AI in public services, digital platforms in organizations, and agile methods in teaching. These evidence notes are helpful when making claims in my text.
  • Idea: I use this node type to capture new ideas that I have not worked on but might have some potential for future research.
  • Theory: This note type collects my references and reflections related to –you guessed it –a specific theory.
  • Manuscript: This is where everything comes together into a coherent text. The beauty of Zettlr is that a manuscript can easily be exported to any format you want, in this way offloading all the formatting tasks to Pandoc so I can concentrate more on my writing.

Thank you for reading! And thanks to all of you who develop and improve Zettlr! Zettlr is an excellent example of how open-source software based on community needs, developed by members of that community, can result in almost perfect products.

References

Ahrens, S. (2017). How to take smart notes: One simple technique to boost writing, learning and thinking. Sönke Ahrens.

Glaser, B. G., & Strauss, A. L. (1967). The discovery of grounded theory: Strategies for qualitative research. Aldine Transaction.

Harris, J. (2017). Rewriting: How to Do Things with Texts (2. Edition). University Press of Colorado.

Leidner, D., & Tona, O. (2021). A Thought-Gear Model of Theorizing from Literature. Journal of the Association for Information Systems, 22(4), 874–892.

Mitchell, K. M., & Clark, A. M. (2021). Enhance Your Qualitative Analysis with Writing: Four Principles of Writing as Inquiry. International Journal of Qualitative Methods, 20.

Watson, R. T., & Webster, J. (2020). Analysing the past to prepare for the future: Writing a literature review a roadmap for release 2.0. Journal of Decision Systems, 29(3), 129–147.